don

Powerflex 755 Net I/O Timeout

8 posts in this topic

I have a CompactLogix system with 4 Powerflex 755 drives and one PanelView.  The drives work (start/stop, speed reference via Ethernet), but fault randomly every 30-120 seconds with a fault 4037 (actually port 4, fault 37, I think) Net I/O Timeout fault.  Fault always clears immediately.  No problems with the Panelview.  Talked to tech support who said it was a network problem and suggested downloading Wireshark, which I did.  I've run it, but not sure what to look for. 

I've disconnected the PanelView and programming computer, but still have the faults.

Network Setup (processor and drives have autonegotiate off, set to 100/full duplex) (all devices have mask 255.255.0.0, gateway 10.91.1.1

10.91.2.5, Processor, 5069-L32ER, firmware 30, using port 2, port 1 is set to DHCP and not connected to anything

10.91.251.5, Switch, N-tron 708TX, IGMP snoping enabled,

10.91.101.5, VFD1, PowerFlex 755 with ENETR card, IP set by parameters

10.91.102.5, VFD2, PowerFlex 755 with ENETR card, IP set by parameters

10.91.103.5, VFD3, PowerFlex 755 with ENETR card, IP set by parameters

10.91.104.5, VFD1, PowerFlex 755 with ENETR card, IP set by parameters

10.91.10.5, PanelView, 2711P-T10C22DP

I've done about 10 similar projects with 4-10 drives and haven't had this problem.  This is the first for this particular processor and using version 30.  Panel is at the panel shop for testing with no motors attached.

Thanks in advance for any ideas.

Share this post


Link to post
Share on other sites
Quote

(processor and drives have autonegotiate off, set to 100/full duplex

Why ?   Unless you have also set the switch to 100/full and disabled auto-negotiate, you might be getting link drops in order to attempt auto-negotiation periodically.

I do not recommend disabling auto-negotiation unless you have hardware (like some fiber/copper converters) that requires it.

Your Class A subnet suggests these drives are on an enterprise network.   Are they isolated, or seeing a lot of other broadcast / multicast traffic ?


Have you set up a port on the 708TX to mirror the CompactLogix traffic to ?   That will be where you connect and use Wireshark.   Generally the 708TX is a battle tank of a switch, well suited to ordinary automation projects.

With the faults being so frequent, you should be able to get some diagnostics just from the link status LEDs on the switch ports to which the PowerFlex are connected.

Are there any other I/O adapters or devices that also show faults/disconnections in the I/O tree of the controller ?  You may need to monitor those with GSV instructions or just watch the status indicators in the I/O tree while online.

 

Share this post


Link to post
Share on other sites
3 hours ago, Ken Roach said:

Why ?   Unless you have also set the switch to 100/full and disabled auto-negotiate, you might be getting link drops in order to attempt auto-negotiation periodically.

I do not recommend disabling auto-negotiation unless you have hardware (like some fiber/copper converters) that requires it.



Your Class A subnet suggests these drives are on an enterprise network.   Are they isolated, or seeing a lot of other broadcast / multicast traffic ?


Have you set up a port on the 708TX to mirror the CompactLogix traffic to ?   That will be where you connect and use Wireshark.   Generally the 708TX is a battle tank of a switch, well suited to ordinary automation projects.

With the faults being so frequent, you should be able to get some diagnostics just from the link status LEDs on the switch ports to which the PowerFlex are connected.

Are there any other I/O adapters or devices that also show faults/disconnections in the I/O tree of the controller ?  You may need to monitor those with GSV instructions or just watch the status indicators in the I/O tree while online.

 

I disabled auto-negotiate because on a previous project, Rockwell tech support recommended do it.  Admittedly, not a very good reason.

I disabled auto-negotiate on the N-tron and that cleared up the problem.  I needed to move and check the rest of the panel and logic.  When I get a lull, I will go back and turn auto-negotiate back on everywhere and see if that works.  I'll also mirror the port and play around with Wireshark some more.

No external network was connected because we were at the panel shop testing the panel.

No other I/O adapters other than the drives.

Thanks for the suggestions.

Share this post


Link to post
Share on other sites

"Auto-Negotiate" is not the same as "Auto-Detect".

A port that's set for auto-negotiation, depending on the exact implementation in the low-level firmware, will periodically drop the link and try to re-negotiate a link speed and duplex, even if it's already running at the highest link speed and full-duplex supported by the device.    

With most ordinary IT gear, that's not a problem.   Your e-mail and web browsing and even YouTube won't notice a 100 ms drop in the link.

But a cyclic I/O connection generally will.

Let us know your outcomes, if you get an opportunity.

 

Share this post


Link to post
Share on other sites

Network issues are never easy. Some questions.. Do the drives only fault when they are running or idle too?  Does the fault randomly change to any of the drives or is it mainly 1? Have you tried isolating a drive or bypassing the switch?

Share this post


Link to post
Share on other sites

Testing  got cut short by CV19. Still no start-up schedule. 

As stated, I got it working in the shop by turning auto negotiation OFF in the N-tron and the drives.   Is it recommended to turn ON auto negotiate in all Rockwell hardware (CPU, drives, etc.) as well as the switch?  I can try that when I get back to it.

Also, I was having the problem no matter how many drives were connected.  Nothing else is on the network, at least during testing at the shop.  Not sure what else the end user is doing with the network, but I assumed only this panel's devices would be on that network.  

 

Share this post


Link to post
Share on other sites

I would look into the switch settings. 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now